Selectable field and frame based predictive video coding

Abstract

An encoder considers a frame representing a picture as comprised of areas. For each area, the encoder decides which of frame-based or field-based orthogonal transformation will be most efficient at reducing spatial redundancy in that area. For each area, the encoder decides which of frame-based or field-based predictive encoding will be most efficient at reducing temporal redundancy in that area. The encoder encodes each area of the picture frame using the most efficient orthogonal transformation technique and using the most efficient predictive encoding technique to produce an encoded signal. A decoder decodes the encoded signal. The encoded signal is recorded on a recording medium, transmitted over a transmission channel, or broadcast.

What is claimed is: 
     
       1. A picture signal decoding method comprising the steps of: receiving an encoded signal including an encoded picture and a header indicating which of frame-based predictive encoding and field-based predictive encoding was used to encode said encoded picture, said encoded picture being encoded with the one of said frame-based predictive encoding and said field-based predictive encoding which produces the lesser amount of data;   frame-based predictively decoding said encoded picture when said header indicates that said encoded picture was encoded using frame-based predictive encoding; and   field-based predictively decoding said encoded picture when said header indicates that said encoded picture was encoded using field-based predictive encoding.   
     
     
       2. A picture signal decoding apparatus comprising: means for receiving an encoded signal including an encoded picture and a header indicating which of frame-based predictive encoding and field-based predictive encoding was used to encode said encoded picture, said encoded picture being encoded with the one of said frame-based predictive encoding and said field-based predictive encoding which produces the lesser amount of data;   means for frame-based predictively decoding said encoded picture when said header indicates that said encoded picture was encoded using frame-based predictive encoding; and   means for field-based predictively decoding said encoded picture when said header indicates that said encoded picture was encoded using field-based predictive encoding.   
     
     
       3. An encoded picture signal produced by: receiving an interlaced signal having frames each containing an odd field and an even field and representing a current picture and at least one other picture;   first evaluating said odd and even fields for said frame representing said current picture to determine a first amount of data which would result from frame-based predictive encoding of said current picture;   second evaluating said odd and even fields for said frame representing said current picture to determine a second amount of data which would result from field-based predictive encoding of said current picture;   comparing said first and second amounts of data;   selecting the one of said frame-based predictive encoding and said field-based predictive encoding which corresponds to the lesser of said first and second amounts of data as determined by the comparison; and   predictively encoding said current picture relative to at least one of the other pictures represented by said interlaced signal using the selected one of said frame-based predictive encoding and said field-based predictive encoding.   
     
     
       4. A picture signal encoding method comprising the steps of: receiving an interlaced signal having frames each containing an odd field and an even field and representing a current picture and at least one other picture;   first evaluating said odd and even fields for said frame representing said current picture to determine a first amount of data which would result from frame-based predictive encoding of said current picture;   second evaluating said odd and even fields for said frame representing said current picture to determine a second amount of data which would result from field-based predictive encoding of said current picture;   comparing said first and second amounts of data;   selecting the one of said frame-based predictive encoding and said field-based predictive encoding which corresponds to the lesser of said first and second amounts of data as determined by the comparison;   predictively encoding said current picture relative to at least one of the other pictures represented by said interlaced signal using the selected one of said frame-based predictive encoding and said field-based predictive encoding; and   appending information to the predictively encoded current picture identifying which of frame-based and field-based predictive encoding was used to encode said predictively encoded current picture.